Do international trade agreements exacerbate climate damage? If so, to what extent? We answer these questions using a general equilibrium framework that accounts for salient features of the global economy, such as input-output linkages, returns to scale, endogenous energy supply, and abatement. We derive simple formulas for the climate cost of trade liberalization following an arbitrary trade agreement. Our formulas decompose trade liberalization's impact on the climate into scale, technique, and composition effects. We map our formulas to data on global production, taxes, carbon emissions, and trade flows to estimate the climate damage attributable to overall globalization trends and specific trade agreements.

We evaluate the impact of the expected International Maritime Organization 2023 regulation (henceforth IMO2023) that caps CO2 emissions from global maritime shipping. Focusing on US imports—where we use detailed vessel, route, emissions, speed, and trade data—we structurally
estimate an import demand model with vessel- and air-delivered imports. We show that IMO2023 will cause demand substitution towards more carbon-intensive air transport to the point that total transport-related CO2 emissions increase. Furthermore, we show that subsidizing fuel for vessels will lower combined air and vessel emissions.

We study deforestation in a dynamic world trade system. Using a data set on global deforestation, we document that between 1990-2020: (i) forest area has decreased by
7.1 percent at the global level, with large heterogeneity across countries, (ii) deforestation has been markedly associated with expansions of agricultural land use, and
(iii) forest carbon intensity differs substantially across the world. We build a model in which structural change and comparative advantage determine the extent, location,
and timing of deforestation. We show analytically that, if agriculture is complementary to the rest of the economy, multilateral agricultural liberalization tends to reduce
deforestation, even if such shocks increase deforestation when experienced only by an individual economy. In our calibrated model, a permanent 30 percent reduction in
global agricultural trade costs leads to a 0.80 percent steady-state increase in global forest area relative to the business as usual scenario—an effect that takes decades to
unfold. Countries that have a comparative advantage in agriculture deforest more to supply agricultural products to the rest of the world. For example, Brazil and Canada
lose 0.2 and 0.4 percent in forest area, while China gains 2.7 percent.

Global production of palm oil has soared in recent years to keep up with growing demand. Demand for palm oil has been growing much faster than other competing oils such as soy, canola, and sunflower due to its low cost, versatility, and high yield per hectare. Production of palm oil is a leading cause of deforestation. With the increase in demand for palm oil, the rate of deforestation to meet production demands has also escalated. The Roundtable of Sustainable Palm Oil (RSPO) was formed to combat the deforestation associated with palm oil and create a sustainable pathway of production. RSPO provides a certification standard and trademark for meeting certain sustainability requirements. Countries have adopted certification at different rates depending on a variety of factors. In this paper we exploit these differences in certification to understand its impact on palm oil trade. Certification is costly but it impacts the demand through altering embodied characteristics of the palm oil. Using both detailed product-level level United Nations Comtrade data and RSPO certification data, we estimate consumers’ valuations of traditional versus certified palm oil. The estimates allow us to quantify the value of the certification program.